1
Running head: LINGUISTIC ABSTRACTION
AIe You Who I Think You Say You AIe?: Effects of Linguistic Abstraction on Applicant
Evaluation Using Third party Descriptions
An Honors Thesis (pSYS 499)
by
Kyle Buck
Thesis Advisor Thomas Holtgraves Signed
Ball State University Muncie Indiana April 2015
Expected Date of Graduation May 2015 -----
..
~t:''-o\\
Cl,.hci.~
_1 r5r 41
-lksL. LINGUISTIC ABSTRACTION
2
L.
'2-Lf8~
24
'?DlS
Abstract
.t. C,3
Although language is well understood as a social construct, little attention has been given to the
effect of semantically constructed meaning on individual perception. When people talk about
themselves or others, subtle variations in linguistic style can influence evaluations of the subject.
The current study was designed to examine differences in evaluation based on the principle of
linguistic abstraction. Semantic language exists on a spectrum between the concrete (what is
physically present) and the abstract (that which exists solely as an ideological concept).
Abstraction influences the extent to which a listener infers a casual attribution about the subject.
Two mock letters of recommendation were prepared that varied as a function of abstract
language. It was hypothesized that ajob candidate who was described more abstractly would be
viewed more positively than one who was described concretely. Results of the study and its
implications for future research are discussed at length.
3
LINGUISTIC ABSTRACTION
Acknowledgements
I would like to thank Thomas Holtgraves for taking the time to supervise and instruct me
for the duration of this project. I also thank Jacob Watson for being a supportive colleague and a
good friend. Finally, I would like to thank the Psychological Science department faculty who
helped me with any technological issues that arose.
LINGUISTIC ABSTRACTION
4
Effects of Linguistic Abstraction on Applicant Evaluation Using Third Party Descriptions
In the field of psycho linguistics, the psychological study of language, it is understood that
language can have a significant impact on psychological processes, behaviors, and attitudes. But
these effects are not isolated to the individual: Language is an inherently social process, designed
for communication among people. It is for this reason that language is tied to social action,
perception, and evaluation (Holtgraves, 2002; Smith & Semin, 2007). Interpersonal research in
psychology has proven valuable to the extent that it reflects prominent social constructs, such as
race relations, gender, and stereotype maintenance. Research on interpersonal relationships has
recently turned toward specific elements in language to explain the maintenance of intergroup
relations (e.g. Maas, Salvi, Arcuri, & Semin, 1989; Semin & Fiedler, 1988; Wigboldus, Semin,
& Spears, 2000). Within this line of research, implicit causality (Brown & Fish, 1983), by which
causal attributions are inferred through linguistic structure, has been a prominent area of study
for over three decades. Yet the scope and impact of this phenomenon remains fairly unknown,
especially when making impressions about other people. For this reason, the current study
focuses on the effects of abstract and concrete language on professional evaluations.
Many facets of our modern world depend on the evaluations of individuals by others; it is
a process by which people are hired, graded, elected, and awarded . Those under scrutiny of
evaluation may try to improve their image or likeability through persuasion. Campaigns and
slogans are often tailored to lead the individual to infer beyond the words themselves.
Psycholinguistic concepts explain the effects of language on cognitive processing. Language use
can either be categorized as abstract, discussing ideas and concepts not physically present, or
concrete, focusing on the physical aspects of the world around us. Moreover, the same person
can be described using either abstract or concrete language. By manipulating the levels of
LINGUISTIC ABSTRACTION
5
linguistic abstraction used in written evaluations, namely letters of recommendation, it is
expected that the individual perceptions of prospective raters will significantly differ.
Linguistic Abstraction
Linguistic abstraction is a form of implicit causality that refers to the extent that an
observer can draw causal inferences from specific parts of speech. The Linguistic Category
Model (LCM; Semin & Fiedler, 1998) operationally defines linguistic abstraction through a
combination of adjectives and verbs (from the most abstract to the least abstract): adjectives,
state verbs, interpretive action verbs CIAYs), and descriptive action verbs (DAYs). Adjectives
are abstract trait descriptions. In causal attributions, traits represent a stable cause of behavior
over time, which cannot be confirmed nor disconfirmed as true (for example, how do we know if
someone is honest if we carmot physically perceive honesty?). State verbs, such as "Jack likes
Jill," suggest that a sentence's object (Jill) is responsible for the emotional action from the
sentence's subject (Jack), and usually stems from some inherent quality of the object. lAYs and
DAYs differ in the objectivity of the action being performed . lAYs are the more abstract of the
two, and imply a nonspecific action that could be subject to interpretation (e.g. encourage,
COmiOlt, succeed). DAYs, being the most concrete category of the LCM, encompass physical
action verbs (e.g. hit, jump, write) that are objective and can be readily witnessed (Semin &
Fiedler, 1988).
The difference between levels of abstractness is two-fold. First, the word categories vary
on the principle of consensus, or the frequency with which a trait or behavior can be found in the
general population (Brown & Fish, 1983). Adjectives and other abstract words provide lower
consensus than action verbs. The reasoning follows that not all individuals can be readily
prescribed a specific trait (as traits are difficult to visualize); action verbs, however, represent
LINGUISTIC ABSTRACTION
6
readily available behaviors that many individuals could be perceived of doing. The second
difference involves the distinctiveness of traits and behaviors. The fact that someone is selfish
may not be as cognitively available as an isolated, selfish action, implying that people find it
easier to envision isolated actions as opposed to overlying abstract concepts. It naturally follows
that high consensus can be paired with high distinctiveness and low consensus with low
distinctiveness (Brown & Fish, 1983).
The perception of abstract language seems to go largely unnoticed at the conscious level.
Individuals are often too cognitively busy to allot the time and resources required to analyze all
aspects of incoming social information (Smith & Collins, 2009). This lack of recognition allows
for the perpetuation of stereotypes and expectancies without requiring thoughtful processing on
the part of the listener, rendering linguistic abstraction a powerful influence in making causal
attributions (e.g. Maass et al., 1989). Indeed, it has been found that abstraction is processed in
.such a way that the individual can interpret meanings that go "beyond the impact of the specific
content or subject" in question (Wigboldus et al., 2000, p. 16). Over the past 30 years,
psycho linguists have discovered several contexts in which linguistic abstraction plays a major
role.
Applications of Linguistic Abstraction
Abstractness is ranked according to the extent that the word's usage suggests the
responsibility of a stable trait as a cause for an individual's actions. In their study of prejudice
and discrimination, Maass et al. (1989) utilized the LCM to develop the concept of Linguistic
Intergroup Bias (LIB), which explains how causal attributions are made between social ingroups
and outgroups. The LIB states that positive ingroup behaviors and negative outgroup behaviors
are described more abstractly to represent the stability over time. Conversely, negative ingroup
LINGUISTIC ABSTRACTION
7
behaviors and positive outgroup behaviors are described more concretely to emphasize
situational factors rather than any individual disposition (Maass et aI., 1989). The LIB is just one
example of how varying levels of linguistic abstraction affect social perceptions and how these
inferences are involved in the creation and maintenance of stereotypes, prejudices, and bases for
discrimination (Maass et aI., 1989; Wigboldus et aI., 2000). Similarly, the Linguistic Expectancy
Bias (LEB) demonstrates how expectant (or stereotype-consistent) behaviors are described with
greater abstraction than unexpected behaviors (Douglas & Sutton, 2003; Karpinski & von
Hippel, 1996; Wigboldus et aI., 2000). Both LIB and LEB are prevalent examples of how
abstract language holds influence in social situations.
The use of linguistic abstraction does not seem to occur at the conscious level (Douglas
& Sutton, 2003). However, research has suggested that linguistic abstraction can be subjected to
motivated manipulation. In a later study concerning LIB, Maas, Ceccarelli, and Rudin (1996)
conducted an experiment that manipulated the level of "threat" that an outgroup posed to the
ingroup participants; accordingly, threats to ingroup identity resulted in greater linguistic
abstraction used when describing outgroup behaviors. Douglas & Sutton (2003) found similar
motivated processing during a study in which linguistic abstraction was moderated by
communication goals. Finally, Menegatti and Rubini (2014) recently discovered that romantic
partners are more likely to use abstract language (adjectives and state verbs) when persuading
their partner to continue their relationship. These examples of self-defense, achievement, and
persuasion suggest that the scope and influence linguistic abstraction are much more flexible
than originally believed, with seemingly generalized applications in various facets of everyday
communication patterns. Therefore, while the use of abstract language appears to be largely
unconscious, it can be primed into motivated use depending on the context.
LINGUISTIC ABSTRACTION
8
Yet abstract language represents only one side of a continuum: concrete language has
also been shown to have significant effects on perception. For exam£le, concrete statements may
be perceived as "more true" than abstract statements (Hansen & Wanke, 2010). This harkens
back to the LIB in that abstract statements are difficult to prove or disprove, while concrete
statements may provide more cognitive closure (Maass et aI., 1989). It has been shown that
concrete anecdotes are weighed more heavily than abstract information when making decisions
(Borgida & Nisbett, 1977). Concrete language is also employed more when terminating an
interpersonal relationship, such as a romantic break up (Menegatti & Rubini, 2014). These
examples represent the necessity for concreteness when making definitive decisions or
permanently ending engagements.
To summarize, abstract linguistic forms, such as adjectives and state verbs, are difficult to
visualize and contest, making them useful for perpetuating status and image. Concrete forms,
such as lAYs and DAYs, inhibit generalization and are easier to visualize, and prove useful
when required to make definitive decisions. The current study seeks to examine the role of
abstraction in decision-making based on individual qualities in the form of evaluations in order
to uncover which of these two linguistic forms holds more weight.
Abstraction and Evaluative Judgment: The Current Study
Many decisions require individuals to evaluate others through the use of a communicative
medium. Yet very little research exists relating linguistic abstraction to individual impression
management. A recent study has shown abstract language to be associated with social power, in
that people who use abstract language are perceived as possessing power (Wakslak, Smith, &
Han,2014). When constructing a social perception about someone else, it would seem that the
subject ' s power relative to the perceiver plays an important role in evaluating information.
LINGUISTIC ABSTRACTION
9
Levels of abstraction have also been found to differ based on need for closure (Webster,
Kruglanski, & Pattison, 1997). It seems that concrete language, too, holds weight when making
decisions. These examples suggest that both abstract and concrete language are mediating
factors in personal judgment, but the pattern or relationship between the two has yet to be
established.
The current study has been designed to build upon practical approaches to linguistic
abstraction by expanding the understanding of these differences in perception. A letter of
recommendation format was used to measure which of the two has a greater observed effect on
evaluative outcomes: an abstract letter emphasizing the employee's personal traits (and in tum
lending credibility to the letter writer for their perceived power), or a concrete depiction of the
employee's behaviors, which should prove more useful in providing closure for the decision­
making process. In general, it is hypothesized that a difference in abstraction will produce
significantly different applicant evaluations. Past research suggests a variety of effects related to
both abstract and concrete language use, which is often dependent upon the context of what is
being said. Specifically, the effects of linguistic abstraction appear dependent upon the subject,
the speaker, and the purpose. Because of these parameters, initial hypothesizing proves difficult
as these factors interact to produce different outcomes in overall perception across different
situations. Therefore, based on the exploratory nature of the current study, the initial hypothesis
remains open due to the under-established pattern of abstract language use across contexts.
Method
Participants
Fifty participants were recruited from the Ball State University subject pool and through Ball
State University e-mail on a voluntary basis, 11 of whom failed to complete the study.
LINGUISTIC ABSTRACTION
10
Participants were aged 18 and older (median age 19). An incentive was given in the form of ~
hour of research credit for the entry level Psychological Science course.
Materials
Job Description. Each participant was shown a job description that outlined the duties
and expectations for an available Human Resource Manager position.
Letters ofrecommendation. Two mock letters of recommendation were prepared. Both
letters contained recommendations by a previous supervisor of the applicant to the Human
Resource Manager position. The letters were tailored so that their overall message was the same,
but the specific elements of language use differed as a function of abstractness. One letter
described the candidate using adjectives (e.g. helpful , supportive, committed),while the other
letter emphasized specific actions of the candidate using IA V s (e.g. helped, supported,
committed). These letters are presented in the Appendix.
Candidate questionnaire. A questionnaire followed the letter of recommendation. The
first set of questions asked participants about different qualities of the employee's confidence
and job fit, the supervisor's confidence and trustworthiness, and hireability. The second set of
items required participants to rate the candidate's basic personal traits (creative, encouraging,
supportive, helpful, committed, organized, productive). A final set of items elicited inferences
from the paJiicipants based on the extent that they agreed with generalized or specific statements
about the candidate's character. These included trait congruent statements, which corresponded
with abstract language, and trait incongruent statements, which corresponded with concrete
language, and thus represent exceptions rather than an overarching trait. The answer format for
all questions followed a 7-point Likert scale, asking respondents, for example, the extent to
LINGUISTIC ABSTRACTION
11
which they agree with the given statements (e.g. 1 = strongly disagree, 7 = strongly agree). The
exact questions and their corresponding Likert scales are presented in the Appendix.
Procedure
Participants accessed the study online through Qualtrics. Participants were told that the
purpose of study was to read and evaluate a letter of recommendation based on the candidate,
writer and job fit. After giving informed consent, each participant was shown the job description
for a Human Resource Manager position, followed by one of the two letters concerning Jim, a
prospective applicant for the position. Each letter was followed by the candidate questionnaire,
to which participants responded to the hireability, trait, and inference scales. The two variants of
the letter shown to the participants are presented in the Appendix.
Results
Measurements
I hypothesized that abstract and concrete language would produce significantly different
applicant evaluations. The three question domains examined were candidate hireability,
candidate traits, and inferences that could be made about the candidate. All of these results are
summarized in Table 1.
Table 1
Independent-Samples t-tests for the Three Domains
Abstract
Concrete
Measure
Mean
SD
Mean
SD
df
1
Hireability
36.69
4.89
36.38
4.04
37
.209
Traits
42.85
5.87
43.31
4.71
35
-.265
Inferences
60.36
6.10
63.11
9.50
37
-.883
12
LINGUISTIC ABSTRACTION
Candidate hireability. The first set of six questions comprised a hireability scale (see
appendix;
0,=
.866). These questions elicited perceptions of candidate fit, confidence of both the
candidate and supervisor, and overall hireablility (e.g. To what extent does Jim appear fit for this
job?) . An independent-samples t-test found no significant difference between abstract and
concrete conditions for hireability (t
(M= 36.7, SD
=
=
.209, P > .1). Participants who read the abstract condition
4.9) did not rate the six items differently than those who read the concrete
condition (M = 36.4, SD = 4.0).
Candidate traits. The second set of seven items acted as a manipulation check to test
whether the abstract and concrete conditions created a difference in perception (a = .849). These
items consisted of a basic trait analysis (Jim is helpfid, Jim is supportive, etc.). An independent­
samples t-test found no significant difference between abstract and concrete conditions for
perceptions of the candidate's traits (t = -.265, P > .1). Participants who read the abstract
condition (M = 42.9, SD
=
5.9) did not rate these items differently than those who read the
concrete condition (M = 43.3 , SD = 4.7).
Inferences about the candidate. The last fourteen items served as a higher-order
manipulation check used to gage whether the participants made different inferences about the
candidate as a person as a function of the abstract and concrete conditions (a = .741). These
items asked the extent to which participants agreed with various statements that either
generalized Jim's behavior (Jim regularly volunteers outside ofwork) or did not (Jim tries to
appear helpfulfor the sake ofhis image). Removal of the item Jim makes an effort to work well
with others raised internal reliability to .778. An independent-samples t-test found no significant
difference between abstract and concrete conditions for the types of inferences made about the
candidate (t= -.883, p > .1). Participants who read the abstract condition (M= 60.4, SD
=
6.1)
LINGUISTIC ABSTRACTION
13
did not rate these items differently than those who read the concrete condition (M = 63.1, SD
=
4.0). Because of the insignificant differences found in each item set and overall, the original
hypothesis cannot be supported.
Exploratory Analysis
[-test for hireability items. Because the overall hireability measure contains
heterogeneous items that measure various aspects of the candidate, supervisor, and hireability,
further independent-samples t-tests were performed for each individual item. These t-tests found
no significant differences between abstract and concrete conditions for individual scores on these
items. These results are summarized in Table 2.
Table 2
Independent-Samples t-tests for Individual Hireability Items
Abstract
Concrete
Question
Mean
SD
Mean
SD
df
t
Supervisor
6.69
.855
6.69
.549
37
.000
6.62
.650
6.42
.703
37
.825
6.15
.987
5.92
.935
37
.714
5.31
1.109
5.46
1.029
37
-.429
6.l5
1.214
6.08
.935
37
.219
confidence
Candidate
confidence
Likelihood of
being hired
Supervisor
trustworthiness
Candidate job
fit
LINGUISTIC ABSTRACTION
Likelihood of
5.77
14
1.166
5.81
1.059
37
-.103
being hired by
participant
General Discussion
Linguistic abstraction affects our perception of language in subtle ways outside of
conscious awareness. Most importantly, linguistic abstraction has been shown to influence what
inferences people can draw about others. For this reason, the current study was designed to
measure differences in professional evaluation based on differences in abstract language use,
namely differences in individual person perception based on judgments in a hiring decision. The
initial hypothesis stated that a significant difference in hireability, trait, and inference ratings
would occur as a function of abstract or concrete language; this was not supported. This study
attempted to find differences in individual person perception based on judgments made about
hypothetical actors in a personally unrelated hiring decision. Although participants were asked
to make judgments based on their perception ofthe material, the influence of linguistic
abstraction failed to manifest when decisions were to be made about a scenario which was both
imaginary, unrelated, and of likely low importance to the individual. This shift in context, then,
represents the point of departure of the current study from previous experiments. However, the
failure ofthe manipulation to find significant differences in perception renders the effects of
linguistic abstraction in this context inconclusive.
In addition to the insignificant t values, it should be noted that the means for each item set
appear quite high when compared to the highest possible score for each set. This suggests that a
ceiling effect has occurred in which responses have become grouped within the upper range of
LINGUISTIC ABSTRACTION
15
possible scores. This ceiling effect helps explain the failure of the manipulation because it is
possible that the scores reflect participant acquiescence rather than an actual effect of linguistic
abstraction.
Implications of the Current Study
The implications and conclusions that can be drawn from the findings remain speculative;
however, the nature of the results inspires a new hypothesis on the contextual limits for the effect
of linguistic abstraction. Previous studies and conceptions such as the LIB model of abstraction
(Maas et al. 1989) demonstrate that this type of language use shows a large effect when the
information being given is personally relevant, such as that used during interracial dialogue or
expressing political partisanship. The LEB model of abstraction (Douglas & Sutton 2003) also
functions based upon individual experience and what is expected to be true. The current study
displaces this psychological distance even further by asking patiicipants to make judgments
about a hypothetical person who is being evaluated on the merit of his work as described by a
supervisor. It could be possible that the power of linguistic abstraction diminishes as the person
or group in question becomes further and further distanced in relation to the individual, to the
point where abstractly- or concretely-charged language has no bearing on the listener's
interpretation. The nature of the current study has led me to consider this hypothesis, yet the
current results render the support for this speculation inconclusive.
Perhaps the most practical suggestion that can be drawn from the current study is related
to the structure of experimental designs in psycho linguistic research. Variations in language use
are very subtle forms of manipulation, but this issue is not restricted to research on linguistic
abstraction. Any time a linguistic concept is hypothesized to have an effect on psychological
. processing, good experimental design dictates that the manipUlation be set up in such a way that
LINGUISTIC ABSTRACTION
16
researchers can say with certainty that any differences in the dependent variable are the result of
the independent variable. However, language itself remains a very intangible concept and
language recognition cannot be observed directly. It should be in the interest of psycholinguistic
researchers to use physical measures to establish operational definitions of their independent
variables. The operational definitions for linguistic abstraction used in the current study were
scores from the measures of hireability, trait recognition, and inference making. The
experimental design may have been strengthened from the use of physical measures as
operational definitions, such as eye tracking, reading times, or electric signals from the brain. I
think that future research on linguistic abstraction can benefit greatly by implementing physical
measures in experimental designs.
Strengths and Limitations
Three factors likely inhibited the success of the current study. First, low participant
turnout and the considerable rate of attrition likely affected the final results. At the departmental
level, this study was being performed in competition for participants for several other, co­
occurring studies, resulting in a smaller sample size. A handful of participants also failed to
complete the study after being exposed to the experimental condition. Within the time frame
allowed for the current study, not enough participants could be recruited to form a substantial
sample size, affecting the overall results. Second, this study was distributed online due to
convenience. Third , it is possible that the manipulation was unable to find significant differences
in perception given the subtle manipulation of language involved in linguistic abstraction.
However, the results also demonstrated one considerably strong feature, namely the high alpha
levels of each scale, indicating that they have the potential to measure their respective constructs
reliably. These scales could prove useful in a future study involving a stronger manipulation.
LINGUISTIC ABSTRACTION
17
Future Research and Conclusions
Future research in this area should consider utilizing a laboratory from which to measure
participant reactions. It is possible that the lack of physical setting inhibited the effects of such
subtle variations in language. Without pragmatic context and experimenter supervision, results
could vary greatly based on the environment in which the participant accessed the survey (time,
location, distractions, etc.). Future research should also consider expanding into factorial designs
measuring interactions between linguistic abstraction and other variables. Regardless of the
method used, further research should focus on increasing statistical power in order to find more
meaningful results (significant or nonsignificant). In general, studies on linguistic abstraction
should focus on utilizing more innovative designs and observable operational definitions to
account for subtle differences in language.
The current study sought to measure differences in perception caused by linguistic
abstraction, yet found no significant results when abstraction was utilized in a decision-making
context (e.g. hiring decisions) . Despite being unable to reject the null hypothesis (that the
conditions would not produce significant differences), the scales used in this study show promise
for future research as measurements of hireability, personal traits, and making inferences. It
remains unclear whether linguistic abstraction mediates judgment when asked to use lexically
charged information to make a decision. However, such a relationship should prove important
when trying to inhibit sources of bias, the kinds of which previous models of linguistic
abstraction have discovered.
LINGUISTIC ABSTRACTION
18
References
Borgida, E., & Nisbett, R. E. (1977). The differential impact of abstract vs. concrete information
on decisions. Journal ofApplied Social Psychology, 7,258-271.
Brown, R., & Fish, D. (1983). The psychological causality implicit in language. Cognition, 14,
237-273.
Douglas, K. M., & Sutton, R. M. (2003). Effects of communications goals and expectancies on
language abstraction. Journal ofPsychology and Social Psychology, 84(4), 682-696.
Hansen, J., & Wanke, M. (2010). Truth from language and truth from fit: The impact of
linguistic concreteness and level of construal on subjective truth. Personality and Social
Psychology Bulletin, 36(11), 1576-1588.
Hoffman, C., & Tchir, M. A. (1990). Interpersonal verbs and dispositional adjectives: The
psychology of causality embodied in language. Journal ofPersonality and Social
Psychology, 58(5), 765-778.
Holtgraves, T. M. (2002). Language as social action: Social psychology and language use.
Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Karpinski, A., & von Hippel, W. (1996). The role of the linguistic intergroup bias in expectancy
maintenance. Social Cognition, 14(2), 141-163.
Maass, A., Salvi, D., Arcuri, L., & Semin, G. (1989). Language use in intergroup contexts: The
linguistic intergroup bias. Journal ofPersonality and Social Psychology, 5 7(6), 981-993.
Maass, A., Ceccarelli, R., & Rudin, S. (1996). Linguistic intergroup bias: Evidence for in-group­
protective motivation. Journal ofPersonality and Social Psychology, 71(3),512-526.
Menegatti, M., & Rubini, M. (2014). Initiating, maintaining, or breaking up? The motivated use
of language abstraction in romantic relationships. Social Psychology, 45(5), 408-420.
LINGUISTIC ABSTRACTION
19
Semin, G. R., & Fiedler, K. (1988). The cognitive fLmctions of linguistic categories in describing
persons: Social cognition and language. Journal ofPersonality and Social Psychology,
54(4),558-568.
Semin, G. R., & de Poot, C. J. (1995). The question-answer paradigm: You might regret not
noticing how a question is worded. Journal ofPersonality and Social Psychology, 73(3),
472-480.
Smith, E. R., & Collins, E. C. (2009). Contextualizing person perception: Distributed social
cognition. Psychological Review, 116(2), 343-364.
Smith, E. R., & Semin, G. R. (2007). Situated social cognition. Current Directions in
Psychological Science, 26(3), 132-135.
Wakslak, C. J., Smith, P. K., & Han, A. (2014). Using abstract language signals power. Journal
ofPersonality and Social Psychology, 107(1),41-55.
Webster, D. M., Kruglanski, A. W., & Pattison, D. A. (1997). Motivated language use in
intergroup contexts: Need-for-closure effects on the linguistic intergroup bias. Journal of
Personality and Social Psychology, 72(5), 1122-1131.
Wigboldus, D. H. J., Semin, G. R., & Spears, R. (2000). How do we communicate stereotypes?
Linguistic biases and inferential consequences. Journal ofPersonality and Social
Psychology, 78(1), 5-18.
20
LINGUISTIC ABSTRACTION
Appendix
Informed consent
In this study you will be asked to evaluate a letter of recommendation for a job
candidate written by a previous supervisor. Then you will be asked questions about
the letter, including details about the candidate, writer, and job fit.
To be eligible to participate in this study, you must be at least 18 years of age.
The study will take between 15 and 20 minutes and you will receive 1fz hour of
research credit for PSYS 100.
There are no foreseeable risks or benefits from participating in this study.
Data will be stored on the researcher's password-protected computer for one year.
Participation in this study is completely voluntary and your responses are entirely
anonymous. You may decide to not participate in this study at any time without
prejudice from the investigator. If you decide to not finish the study, you will still be
given research credit.
For questions about your rights as a research subject, please contact Office of
Research Integrity, Ball State University, Muncie, IN 47306, (765) 285-5070,
irb@bsu.edu. For questions about this research, you may contact the principle
investigator; his contact information is provided below:
Kyle Buck
Department of Psychological Science
Ball State University
kabuck2@bsu.edu
Faculty supervisor:
Thomas Holtgraves, Ph.D.
Department of Psychological Science
Ball State University
Muncie, IN 47306
OOtOholtgrav@bsu.edu
765-285-1716
Study overview
You will be shown a job description for the position of Human
Resource Manager, followed by a letter of recommendation written by
a previous supervisor for a potential job candidate. After reading
LfNGUISTIC ABSTRACTION
21
both pieces, you will be asked to answer questions regarding the
letter, including details about the candidate, writer, and job fit.
Job description
Human Resource Managers are expected to organize and oversee workplace
environment. Managers address worker productivity, adjustment, and concerns. They are
responsible for coordinating workplace layout and office assignments. Managers are
expected to be capable of resolving any quality-of-life issues employees have or be able to
provide the resources to resolve these issues. Other responsibilities include:
•
Lead human resource team and oversee human resource projects
•
•
Relay employee and quality-of-life assessments to ChiefExecutive Officer
•
•
Appropriately document and store these assessments in a timely manner
•
•
Provide authority in resolving employee and workplace related conflicts
•
•
Communicate well and consistently with human resource team and other employees
Abstract transcript
Please read thefollowing letter ofrecommendation for Jim Thorpe, who is applyingfor a
Human Resource position. This letter is written by Jim's former supervisor. This letter will be
followed by a series ofquestions related to the content ofthe letter, so please read carefully.
To Whom It May Concern:
Jim Thorpe has been a valuable member of the Human Resource
team at Outreach Services. I have been his supervisor for the past two
years. Jim consistently impresses his peers, our employees, and myself and
would prove a valuable asset in any human resource position.
Much can be said of his work ethic: Jim is a productive individual who
understands what is expected of him. Even under pressure, Jim remains
committed to his work and meets all requirements for any project
admirably. Time and again, Jim was an encouraging member of our
team. There can be little doubt that he is supportive of other coworkers
and exhibits a cooperative attitude, fitting well into any group
composition. He is organized and able to see optimal solutions to problems
LINGUISTIC ABSTRACTION
22
that arise in the structure of the workplace. Jim is at his most creative
when managing employee happiness, and prefers strategic outings and
group-building exercises to foster cohesion among workers. Overall, he
was a helpful person during our company's time of need.
I think that Mr. Thorpe would make a great addition to your company. I
confidently recommend Jim Thorpe for your company's position of Human
Resource Manager.
Concrete Transcript
Please read the following letter ofrecommendation for jim Thorpe, who is applying for a
Human Resource position. This letter is written by jim's former supervisor. This letter will be
followed by a series ofquestions related to the content ofthe letter, so please read carefully.
To Whom It May Concern:
Jim Thorpe has been a valuable member of the Human Resource
team at Outreach Services. I have been his supervisor for the past two
years. Jim consistently impresses his peers, our employees, and
myself and would prove a valuable asset in any human resource
position.
Much can be said of his work ethic: Jim produced work according to
what was expected of him. Even under pressure, Jim was committed
to his work and met all requirements for any project admirably. Time
and again, Jim encouraged the members of our team. There can be
little doubt that he supported other coworkers and cooperated well
with any group composition. He organized employees and was able to
see optimal solutions to problems that arose in the structure of the
workplace. Jim created employee happiness by planning strategic
outings and group-building exercises to foster cohesion among
workers. He has always helped us when we needed it most.
I think that Mr. Thorpe would make a great addition to your
company. I confidently recommend Jim Thorpe for your company's
position of Human Resource Manager.
Scale 1: Candidate hirabilitv
LINGUISTIC ABSTRACTION
23
How confident is Jim's supervisor in his recommendation? (1
confidant)
= Not at all confidant, 7 =very
How confident does Jim appear as an employee? (1 = Not at all confidant, 7 = Very
confidant)
How likely is it that a potential employer would be willing to hire Jim based on this
recommendation? (1 = Not at all likely, 7 =Very likely) To what extent do you trust the supervisor's judgment (who wrote the letter)? (1 = Do not trust, 7 = Trust greatly) To what extent does Jim appear fit for this job? (1 = Not at all fit, 7
= Very fit)
How likely would be to hire Jim based on this recommendation? (1
Very likely) = Not at all likely, 7 = Scale 2: Candidate traits How well do the following statements describe Jim? (1
= Not at all, 7 =Very well) Jim is creative.
Jim is encouraging.
Jim is supportive. Jim is helpful. Jim is committed. Jim is organized. Jim is productive.
Scale 3: Inferences about the candidate
Consider Jim as a person. Based on this letter, to what extent do you agree with the
following statements? (1 = Strongly agree, 7 = Strongly disagree)
Jim performed a majority of the work for the Human Resource team at Outreach
Services.
Jim is only creative to the extent that a practical solution is needed. (R)
LINGUISTIC ABSTRACTION
24
Jim only supports people who share his interests. (R)
Jim commits himself to his significant other.
Jim is only organized when his job demands it. (R)
Jim motivates others by his personality alone.
Jim regularly volunteers outside of work.
Jim takes vacations sparingly
Jim makes an effort to work well with others. (R)
Jim is naturally compelled to support others.
Jim doesn't cooperate
if he doesn't have to.
(R)
Jim tries to appear helpful for the sake ofhis image. (R)
Jim is likely to come up with solutions to many everyday problems.
Jim is only productive while at his job. (R)
Additional questions
What is your age?
If you are enrolled in PSYSCI00, please indicate your first and last name (for the purposes
of receiving course credit). If not, leave blank and proceed.
Debriefing
Thank you for participating in this study! The current study was designed to gauge
people's evaluations of others based on the presentation of a written description. It
is hypothesized that the use of different classes of verbs and adjectives can impact
the overall evaluation of an individual. By studying the connection between language
and inferences, psychologists can begin to recognize patterns of evaluation that arise
from spoken and written communication.
B A L
LM-TAT E
U N I VERSITY
Office of Research Integrity
Institutional Review Board (IRB)
2000 University Avenue
Muncie, IN 47306-0155
Phone: 765-285-5070
DATE:
December 11 , 2014
TO:
Kyle Buck
FROM:
Ball State University IRB
RE:
TITLE:
IRB protocol # 694228-1
Effects of Abstract Language on Applicant Evaluation
New Project
SUBMISSION TYPE:
ACTION:
DECISION DATE:
REVIEW TYPE:
APPROVED
December 11 , 2014
EXEMPT
The Institutional Review Board reviewed your protocol on December 11, 2014 and has determined the
procedures you have proposed are appropriate for exemption under the federal regulations. As such,
there will be no further review of your protocol , and you are cleared to proceed with the procedures
outlined in your protocol. As an exempt study, there is no requirement for continuing review. Your protocol
will remain on file with the IRB as a matter of record .
Exempt Categories:
Category 1: Research conducted in established or commonly accepted educational settings,
involving normal educations practices, such as (i) research on regular and special education
instructional strategies , or (ii) research on the effectiveness of or the comparison among
instructional techniques, curricula, or classroom management methods.
X
Category 2: Research involving the use of educational test (cognitive , diagnostic, aptitude,
achievement), survey procedures, interview procedures or observation of public behavior
Category 3: Research involving the use of educational test (cognitive, diagnostic, aptitude,
achievement), survey procedures, interview procedures, or observation of public behavior
that is not exempt under category 2, if: (i) the human subjects are elected or appointed
officials or candidates for public office; or (ii) Federal statute(s) require(s) without exception
that the confidentiality of the personally identifiable information will be maintained throughout
the research and thereafter.
Category 4: Research involving the collection of study of existing data, documents, records ,
pathological specimens, or diagnostic specimens, if these sources are publicly available or
if the information is recorded by the investigator in such a manner that subjects cannot be
identified , directly or through identifiers linked to the subjects.
- 1-
Gellerat90 Or: IPS! I
Category 5: Research and demonstration projects which are conducted by or subject to
the approval of Department or agency heads , and which are designed to study, evaluate
or otherwise examine: (i) public benefit or service programs ; (ii) procedures for obtaining
benefits or services under those programs ; (iii) possible changes in methods or levels of
payment for benefits or services under these programs.
Category 6: Taste and food quality evaluation and consumer acceptance studies , (i) if
wholesome foods without additives are consumed or (ii) if a food is consumed which contains
a food ingredient at or below the level and for a use found to be safe, by the Food and Drug
Administration or approved by the Environmental Protection Agency or the Food Safety and
Inspection Service of the U.S . Department of Agriculture .
Editorial Notes:
1. Approved- Exempt
While your project does not require continuing review, it is the responsibility of the P.I. (and , if applicable,
faculty supervisor) to inform the IRB if the procedures presented in this protocol are to be modified or if
problems related to human research participants arise in connection with this project. Any procedural
modifications must be evaluated by the IRS before being implemented, as some modifications
may change the review status of this project. Please contact (OR I Staff) if you are unsure whether
your proposed modification requires review or have any questions. Proposed modifications should be
addressed in writing and submitted electronically to the IRB (http://www.bsu.edu/irb) for review. Please
reference the above IRB protocol number in any communication to the IRB regarding this project.
Reminder: Even though your study is exempt from the relevant federal regulations of the Common Rule
(45 CFR 46, subpart A) , you and your research team are not exempt from ethical research practices and
should therefore employ all protections for your participants and their data which are appropriate to your
project.
(r
\...
Bryan Byers, PhD/Chair
Institutional Review Board
Christopher Mangelii , JD, MS , MEd, CIP/Director
Office of Research Integrity
-2­